1. Importing the libraries:¶

In [1]:
import pandas as pd
import numpy as np
import folium
import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import plotly.io as pio
pio.templates.default = 'plotly_white'

import warnings

warnings.filterwarnings("ignore")

2. Reading the dataset:¶

In [2]:
metro_data = pd.read_csv("Delhi-Metro-Network.csv")
In [3]:
metro_data.head(2)
Out[3]:
Station ID Station Name Distance from Start (km) Line Opening Date Station Layout Latitude Longitude
0 1 Jhil Mil 10.3 Red line 2008-04-06 Elevated 28.67579 77.31239
1 2 Welcome [Conn: Red] 46.8 Pink line 2018-10-31 Elevated 28.67180 77.27756

3. Performing EDA:¶

In [4]:
metro_data.describe()
Out[4]:
Station ID Distance from Start (km) Latitude Longitude
count 285.000000 285.000000 285.000000 285.000000
mean 143.000000 19.218947 28.595428 77.029315
std 82.416625 14.002862 0.091316 2.875400
min 1.000000 0.000000 27.920862 28.698807
25% 72.000000 7.300000 28.545828 77.107130
50% 143.000000 17.400000 28.613453 77.207220
75% 214.000000 28.800000 28.666360 77.281165
max 285.000000 52.700000 28.878965 77.554479

3.1 Checking the NULL Values:¶

In [5]:
metro_data.isna().sum()
Out[5]:
Station ID                  0
Station Name                0
Distance from Start (km)    0
Line                        0
Opening Date                0
Station Layout              0
Latitude                    0
Longitude                   0
dtype: int64

3.2 Checking the data types:¶

In [6]:
metro_data.dtypes
Out[6]:
Station ID                    int64
Station Name                 object
Distance from Start (km)    float64
Line                         object
Opening Date                 object
Station Layout               object
Latitude                    float64
Longitude                   float64
dtype: object
The opening date column is represented as an object, so let's convert it to datetime object:¶
In [7]:
metro_data.head(2)
Out[7]:
Station ID Station Name Distance from Start (km) Line Opening Date Station Layout Latitude Longitude
0 1 Jhil Mil 10.3 Red line 2008-04-06 Elevated 28.67579 77.31239
1 2 Welcome [Conn: Red] 46.8 Pink line 2018-10-31 Elevated 28.67180 77.27756
In [8]:
metro_data['Opening Date'] = pd.to_datetime(metro_data['Opening Date'])
In [9]:
metro_data.dtypes
Out[9]:
Station ID                           int64
Station Name                        object
Distance from Start (km)           float64
Line                                object
Opening Date                datetime64[ns]
Station Layout                      object
Latitude                           float64
Longitude                          float64
dtype: object

There are no any NULL values, and the data type for Opening Date column has been changed to datetime format, now, let's start the analysis.

4. Geo-Spatial Analysis:¶

In [10]:
metro_data['Line'].value_counts()
Out[10]:
Blue line            49
Pink line            38
Yellow line          37
Voilet line          34
Red line             29
Magenta line         25
Aqua line            21
Green line           21
Rapid Metro          11
Blue line branch      8
Orange line           6
Gray line             3
Green line branch     3
Name: Line, dtype: int64
In [11]:
metro_data['Line'].value_counts().sum()
Out[11]:
285
In [12]:
metro_data.shape
Out[12]:
(285, 8)
In [13]:
print(f"In total we have: {metro_data['Line'].nunique()} unique lines, " 
      "so, let's assign a unique color to them so that we can plot it.")

In total we have: 13 unique lines, so, let's assign a unique color to them so that we can plot it.

4.1 Assigning unique colors to each of the railway lines:¶

In [14]:
line_colors = {
    "Blue line": "blue",
    "Pink line": "pink",
    "Yellow line": "yellow",
    "Voilet line": "purple",
    "Red line": "red",
    "Magenta line": "black",
    "Aqua line": "lightblue",
    "Green line": "green",
    "Rapid Metro": "cadetblue",
    "Blue line branch": "darkpurple",
    "Orange line": "orange",
    "Gray line": "beige",
    "Green line branch": "lightgreen"
}

Let's visualize the tracks now:

In [15]:
print("The map below shows the line connection of different railway lines, hover over to get more info")
delhi_map_w_line_tooltip = folium.Map(location=[28.7041, 77.1025], zoom_start=11)
for index, row  in metro_data.iterrows():
    line = row['Line']
    color = line_colors.get(line, 'black')  # If line name is not found in dictionary, the default is black
    folium.Marker(
        location=[row['Latitude'], row['Longitude']],
        popup = row['Station Name'],
        tooltip=f" Station name: {row['Station Name']} \n Line: {line}",
        icon=folium.Icon(color=color)
    ).add_to(delhi_map_w_line_tooltip)

delhi_map_w_line_tooltip

The map below shows the line connection of different railway lines, hover over to get more info
Out[15]:
Make this Notebook Trusted to load map: File -> Trust Notebook

5. The growth of railways over the time:¶

5.1 Adding an extra column for the years:¶

In [16]:
metro_data['year'] = metro_data['Opening Date'].dt.year
In [17]:
metro_data.head(2)
Out[17]:
Station ID Station Name Distance from Start (km) Line Opening Date Station Layout Latitude Longitude year
0 1 Jhil Mil 10.3 Red line 2008-04-06 Elevated 28.67579 77.31239 2008
1 2 Welcome [Conn: Red] 46.8 Pink line 2018-10-31 Elevated 28.67180 77.27756 2018
In [18]:
num_of_stations_per_year = metro_data['year'].value_counts()
num_of_stations_per_year_df = num_of_stations_per_year.reset_index()
num_of_stations_per_year_df.head(2)
Out[18]:
index year
0 2018 64
1 2010 54

5.2 Renaming the columns in the new railway count dataframe:¶

In [19]:
num_of_stations_per_year_df.rename(columns={
    "index": "Year",
    "year": "Number of railway tracks"
}, inplace=True)
In [20]:
fig = px.bar(data_frame=num_of_stations_per_year_df, x='Year', y='Number of railway tracks', 
            text='Number of railway tracks',
            title="Number of metro stations opened each year from 2002 to 2019")

fig.update_layout(yaxis_title = "Number of stations",
                 xaxis_tickangle=-45, xaxis=dict(tickmode='linear'))

fig.update_traces(textposition='outside')
fig

Some of the years didn't have any new connection, it could be due to various reasons:

  1. Lack of planning.
  2. Financial fundings or
  3. Construction challenges

From the above chart, 2018 has the highest number of railway tracks opened. Now, let's see which and how many tracks were opened in that year.

In [21]:
metro_2018_data = metro_data[metro_data['year'] == 2018][['Station Name', 'Distance from Start (km)', 'Station Layout']]
In [22]:
print(f"In the year 2018, a total of {metro_2018_data['Distance from Start (km)'].sum():.2f} KM distance were covered "
     "with the addition of new railway tracks.")

In the year 2018, a total of 1540.90 KM distance were covered with the addition of new railway tracks.
In [23]:
station_layoout = metro_2018_data['Station Layout'].value_counts().reset_index()
In [24]:
station_layoout.rename(columns={
    "index": "Station Layout",
    "Station Layout": "Total"
}, inplace=True)
In [25]:
fig = px.bar(data_frame=station_layoout, x="Station Layout", y="Total", color="Station Layout", 
       text="Total", title="Total of different layouts added in the year 2018")

fig.update_traces(textposition='outside', textfont_size=14)
fig

6. Line Analysis:¶

Now, let's analyize the number of lines in terms of the number of stations and the average distance between the stations.

In [26]:
metro_data.head()
Out[26]:
Station ID Station Name Distance from Start (km) Line Opening Date Station Layout Latitude Longitude year
0 1 Jhil Mil 10.3 Red line 2008-04-06 Elevated 28.675790 77.312390 2008
1 2 Welcome [Conn: Red] 46.8 Pink line 2018-10-31 Elevated 28.671800 77.277560 2018
2 3 DLF Phase 3 10.0 Rapid Metro 2013-11-14 Elevated 28.493600 77.093500 2013
3 4 Okhla NSIC 23.8 Magenta line 2017-12-25 Elevated 28.554483 77.264849 2017
4 5 Dwarka Mor 10.2 Blue line 2005-12-30 Elevated 28.619320 77.033260 2005
In [27]:
total_distance_per_line = metro_data.groupby("Line")['Distance from Start (km)'].max().reset_index()
total_number_of_lines = metro_data['Line'].value_counts().reset_index()
In [28]:
total_number_of_lines.rename(columns={'index': 'Line', 'Line':'total_stops'}, inplace=True)
In [29]:
distance_per_line_df = pd.merge(total_distance_per_line, total_number_of_lines, on='Line', how='inner')
In [30]:
distance_per_line_df['average_km_per_station'] = distance_per_line_df['Distance from Start (km)'] / (distance_per_line_df['total_stops'] - 1)
The summary of total distance covered,total number of stops,and average distance in KM per station:¶
In [31]:
distance_per_line_df
Out[31]:
Line Distance from Start (km) total_stops average_km_per_station
0 Aqua line 27.1 21 1.355000
1 Blue line 52.7 49 1.097917
2 Blue line branch 8.1 8 1.157143
3 Gray line 3.9 3 1.950000
4 Green line 24.8 21 1.240000
5 Green line branch 2.1 3 1.050000
6 Magenta line 33.1 25 1.379167
7 Orange line 20.8 6 4.160000
8 Pink line 52.6 38 1.421622
9 Rapid Metro 10.0 11 1.000000
10 Red line 32.7 29 1.167857
11 Voilet line 43.5 34 1.318182
12 Yellow line 45.7 37 1.269444

Now, let's visualize these results:¶

In [32]:
# Creating the sub plots first with 1*2 dimension:
fig = make_subplots(rows=1, cols=2, subplot_titles=("Average distance in KM per stops", "Total number of stops per line"),
                   horizontal_spacing=0.2)

# Plotting the average distance in KM per stops:
fig.add_trace(go.Bar(y=distance_per_line_df['Line'], x=distance_per_line_df['average_km_per_station'],
                    name='Average distance in KM per station', marker_color='crimson', orientation='h'),
             row=1, col=1)



# Plotting the count of stops in each lines:
fig.add_trace(go.Bar(x=distance_per_line_df['total_stops'], y=distance_per_line_df['Line'],
                    name='Tota Number of stops in Each Line', marker_color='navy',orientation='h'),
             row=1, col=2)


# Updating X and Y axis properties:
fig.update_xaxes(title_text="Average Distance in KM", row=1, col=1)
fig.update_xaxes(title_text="Number of stations", row=1,col=2)

fig.update_yaxes(title_text="Metro Lines", row=1, col=1)

# Update layout:
fig.update_layout(height=550, width=950, title_text="Metro Line Analysis")

fig

7. Station Layout Analysis:¶

In [33]:
station_layout_df = metro_data['Station Layout'].value_counts().reset_index()
station_layout_df
Out[33]:
index Station Layout
0 Elevated 214
1 Underground 68
2 At-Grade 3
In [34]:
fig = px.bar(data_frame=station_layout_df, x='index', y='Station Layout', color='index',
        title='Delhi Metro Station Layout Count', text='Station Layout',
        labels={'index':'Layouts', 'Station Layout':'Number of stations'})


fig.update_traces(textposition = 'outside')


fig
In [ ]: